Re: [PATCH v3 08/16] mm/huge_memory: refactor copy_huge_pmd() non-present logic

[Vlastimil Babka]

On 11/10/25 23:21, Lorenzo Stoakes wrote:
> Right now we are inconsistent in our use of thp_migration_supported():
> 
> static inline bool thp_migration_supported(void)
> {
> 	return IS_ENABLED(CONFIG_ARCH_ENABLE_THP_MIGRATION);
> }
> 
> And simply having arbitrary and ugly #ifdef
> CONFIG_ARCH_ENABLE_THP_MIGRATION blocks in code.
> 
> This is exhibited in copy_huge_pmd(), which inserts a large #ifdef
> CONFIG_ARCH_ENABLE_THP_MIGRATION block and an if-branch which is difficult
> to follow
> 
> It's difficult to follow the logic of such a large function and the
> non-present PMD logic is clearly separate as it sits in a giant if-branch.
> 
> Therefore this patch both separates out the logic and utilises
> thp_migration_supported().
> 
> No functional change intended.
> 
> Signed-off-by: Lorenzo Stoakes <lorenzo.stoakes@xxxxxxxxxx>

Yeah much better.

Reviewed-by: Vlastimil Babka <vbabka@xxxxxxx>